What’s the Pattern for Senate Seats in Mid-Terms?
Looking at past Senate elections and some more thoughts on historical patterns.
A couple of days ago I looked at the pattern of seat loss (and rare gain) by the president’s party in the House during the mid-terms (which is usually where these discussions focus). But what about the Senate?
It looks likely that the Democrats will end up with +1 (or, perhaps, +0).
So how does that figure into the historical mix?
Returning to UCSB’s American Presidency Project, we find the following:
Note that in the corresponding House races, the president’s party lost seats, except in 2002 (with an average loss, not counting 2002, of 30.9 seats).
So, in terms of historical context, the Senate outcome on this metric is not outside historical patterns in the least.
Of course, Senate contests are very different given that only one-third of the chamber is up for reelection in a given cycle and the partisan make-up of the states in question can very much influence the possibilities.
It is worth noting that the Democratic losses in 2010 were not only part of a general anti-Democratic wave (we were still dealing with the Great Recession and the party lost 63 House seats) it is worth noting that 2008 was an especially good year for the Democratic Party (including a brief filibuster-proof majority in the Senate). The 2010 mid-terms were a great example of the bigger they are the harder they fall.
Also noteworthy: 1994 was the Republican Revolution.
At any rate, any claims about the Senate contests being of “historic” proportions would be incorrect. This outcome, whether is it +1 or 0, is well within expected norms.
The “historic” issue is going to be based on the House. It would seem that at a minimum the president’s party will lose fewer seats than the pattern suggests and there is still the outside chance that Biden’s party will retain control of the Chamber, which would put 2022 in the “historical” column–although not as anomalous at FDR and Bush in gaining seats at the first mid-term (which I think is still mathematically possible but seems especially unlikely). The Democrats have 222 seats in the current Congress and 218 are needed for majority control of the next (so they can both lose seats, fitting the normal pattern, and still retain control). (Again, I will remind everyone that the map upon which the House candidates are competing was always going to be close, barring something highly unusual).
Historical side-note: Clinton was the first president to have his party lose control of the House since Truman in 1946, since Democrats controlled the chamber from 1955 to 1995. So while JFK, LBJ, and Carter all lost seats in the House in their mid-terms, they did not lose control of the House.